As a submission for the Data Science for Good: City of Los Angeles Kaggle competition, I introduce a high-precision program that automatically visualizes the promotional pathways available for current city employees with only raw text job postings. I use this pathway data, LA job applicant diversity figures, and t-SNE dimensionality reduction to assess sources of bias that could discourage a diverse and talented new generation of civil servants in Los Angeles.
For a full explanation of the results refer to the Kaggle kernel, which won 5th place out of 115, or fork the repo to replicate.